Amazon Web Services
Amazon Web Services is Amazon’s cloud platform: servers, storage, databases and networks switched on as you use them, with no hardware to buy and no machine room to run. It is a sensible choice when workload grows in bursts, when an environment needs replicating in another geographic area or when the server renewal cycle has become a cash problem. Aesir sizes, configures and manages it on the company’s behalf.
Amazon Web Services at a glance
Pay-as-you-go resources
Virtual machines, disk space and databases are switched on when needed and off when no longer needed. You pay for usage time and data volume, not a fixed commitment decided three years earlier. Sizing is no longer a bet on the peak, but a decision that can be corrected.
Regions and availability zones
AWS divides the world into geographic regions, and each region into several availability zones: separate data centres, independently powered and connected. Spreading an application across two zones in the same region is the standard way to survive the failure of a single site without building your own second data centre.
Managed services
Beyond virtual machines, AWS offers ready-made components: relational databases run by the vendor, message queues, load balancers, serverless functions. The benefit is removing repetitive systems work – patches, backups, failover – where it adds no value. The drawback is a tighter tie to the platform, a choice to make consciously.
Identity and permissions
Every access, human or application, goes through an identity and policy system. You define who can do what on which resource, separate production from test environments and track administrative actions. It is the part most often neglected in a cloud opened in a hurry, and the one that determines how much damage a stolen credential can do.
Where data resides
AWS has European regions, Milan included, and the customer chooses which region its data resides in and from which regions it can be reached. The choice should be written down and verified: not all services are available everywhere, and some ancillary functions may involve transfers. We go deeper into the topic on the page Digital sovereignty.
Spending control
The cloud bill grows by itself if nobody watches it: test environments left running, orphaned disks, historical data on fast storage. Resource tagging, budgets with alert thresholds and periodic sizing reviews are part of the service, not an annual check.
Amazon Web Services features
The core of the platform is a few building blocks that combine: compute, storage, private network, identity. On top of these AWS builds more specialised services – managed databases, data analytics, containers, event-driven functions. For most SMEs the value lies in the first two layers: moving application servers and data to the cloud, with a well-designed network, verified backups and a second recovery site. Advanced features make sense when there is a concrete problem that requires them.
On jurisdiction it pays to be precise, because it is a legitimate and frequent question. Data can be hosted in European regions and stay there; the provider, however, is a US company and as such is subject to US law, including the CLOUD Act, which allows US authorities to request data held by a company under their jurisdiction even when stored abroad. This does not make the service unusable or unlawful: there is a data transfer framework between the European Union and the United States, AWS publishes contractual commitments on data location and on handling authority requests, and it is developing a European offering with separate governance. It does mean the choice must be made knowingly, weighing the sensitivity of the data involved and, where needed, encrypting it with keys that remain under the customer’s control. We discuss this openly in Digital sovereignty.
- Virtual machines (EC2): Linux and Windows servers of various sizes, resizable over time. They are the most direct way to move existing applications to the cloud without rewriting them.
- Object storage (S3): low-cost storage of files and objects, with different classes for hot and historical data. Suited to backups, document archives and application content.
- Managed databases (RDS): relational engines such as PostgreSQL, MySQL and SQL Server with backups, updates and replication run by the platform instead of the sysadmin.
- Virtual private network (VPC): subnets, traffic rules and encrypted links to the office. The cloud is not an appendix exposed to the Internet, but an extension of the company network.
- Identity and access (IAM): users, roles and policies following least privilege, multi-factor authentication on admin accounts and separation between environments.
- Auto scaling and load balancing: the number of instances follows the load and traffic is spread across active machines.
- Backup and restore: scheduled copies of machines and databases, kept under defined retention rules and tested with real restores, not just declared.
- Monitoring and logs (CloudWatch, CloudTrail): metrics, alerts and a trail of admin actions. It helps spot problems before users do and reconstruct what happened.
- Region selection: the geographic location of data is a design parameter, declared to the customer and verifiable, not a random consequence of the initial setup.
How we put it into operation
We start from the process, not the platform. We look at which applications support daily work, how much downtime they tolerate, how much data they move and which constraints they have – contractual, regulatory or integration with systems that stay on site. Sizing comes from there: which workloads make sense to move, at what size, in which region, with which backup and recovery scheme. We also say what is better not to move: public cloud is not the right answer to everything, and a project starting from an honest analysis costs less than one starting from an optimistic quote.
Then we implement and migrate, usually in phases, with an agreed window and the option to roll back until the new environment is stable. After go-live comes the longest part: monitoring, updates, restore testing, periodic review of spending and sizing, user support. The customer has one contractual point of contact for platform, network and workstations, instead of having to referee between suppliers when something does not work.
Amazon Web Services is part of Cloud and Infrastructure. See also the Aesir method and our success stories.
Want to know whether Amazon Web Services makes sense for your company?
We start from the process, not the product. A first meeting with no commitment.